Broad objectives. The main objective is to describe and explain the relationship between age and changes in the sense of control over one's life. Specific aims. Two cross-sectional pilot surveys (One Illinois, one U.S.) show that the average sense of control declines at an accelerating rate in successively older age groups. The proposed study will measure changes in the sense of control over a follow-up period, correlate the changes with age, and test hypotheses about the form and components of the association. The main hypotheses are: (I)over a period of time, the sense of control declines by an amount that increases with age; (II) the change in sense of control reflects an underlying change in biosocial function, which declines more rapidly with age; (III) higher social status slows the decline in the sense of control, possibly by preserving biosocial function; (IV) the lower the sense of control at the beginning of a period, the more it drops over the period, so the decrease gets steeper with age, and (V) changes in biosocial function and in the sense of control have deviation-amplifying reciprocal effects that accelerate age- dependent changes in the sense of control. Health relatedness. If Rodin's theory of aging dynamics is correct, a sense of control over one's life increases active participation and improves emotional well-being and physical function, which in turn bolsters the sense of control. Emotional well-being and physical function are health factors themselves, and they decrease subsequent morbidity and mortality. Study design and methods. The study will survey a national sample of persons ages 18 and over. Respondents will be selected at random, interviewed by telephone, and interviewed again three years later. The probability of being sampled will increase with age, to ensure an adequate sample of older persons. The analysis will specify models that test the hypotheses and address technical issues (correlated error, non-random attrition, and interactions or polynomials in structural equation models).